https://www.duolingo.com/zeande

Duolingo Sentence Parsing

This is more-so a question of the implementation of Duolingo after it takes translated sentences (so hopefully I can get a Duolingo dev to answer this). In particular, Duolingo sometimes incorrectly parses content from articles (it seems like it does its parsing based on periods except after ordinal numbers). So if there was, for example, an abbreviation of some sort, it often gets split up into two sentences.

When this happens, for instance say one German sentence in an article has been parsed into two different pieces, should we look at both parts of the one sentence and put the whole translation into both input fields or should we also cut the translation into two parts and put each in their corresponding input fields?

The main reason I ask is because when the English translations are being implemented on the web, there could be problems depending on how sentences are put together. What is everyone else doing when they come across such sentences?

EDIT: A good example of this is the article "Johannes Mario Simmel" sentences 5 and 6.

April 23, 2012

0 Comments

Learn a language in just 5 minutes a day. For free.